Good advices for a newbie (threading tools lathe)

[rendoman]

Thanks for advice!

I did some calculation and attempts in order to understand better the trouble...
In summary:

-initial circuit run at 55v between gnd and cathode, with 220ohm cathode resistor are 250ma current, 3 times and something over normal operation. Not good, this working point destroyed first resistor

- I added a 2,7k resistor from power supply to g2 of el34, a little better situation, but Always too much current. 37.5v gnd-K, 170ma anode current. (normal working point "should" be 300v, 83ma with -15v Vg)

Components are new and show correct measure with multimeter.
There remains only the anode choke, maybe the tube doesn't appreciate so much. Maybe turns of copper are not enough, and inductance need to be really higher, like in commercial product.
I got 2 way, for what I see:

- recalculate a coil with 100H of primary inductance. Lot of copper, this time I use lathe
- Try a high power alu resistor as you said, maybe 50w or more, and throw to the air some 25w. This way allows me to raise B+ for this circuit to 500v.

...everyday a new problem

[rendoman]

New update

Second bad evening, due to circuit.

I was losing my head to look for the cause of the problem, last time I lost some capacitors and resistors.
I found that one tube Langrex, "matched" quad, was faulty. In the same circuit, with the other 3 tube of the same lot everything is ok.
Bad tube runs at 35v of bias, +/- 160ma, current so high that powersupply suffered.

A couple of number:

- Normal value: B+ 300v , 290v on g2, 16v between cathode and ground, with 220ohm are 73ma of current. Good working point

- Bad tube: B+ 270v , 233v on g2 , 35v between cathode and ground, with 220 ohm are 159ma of current. This combo melt cathode resistor.

I modified also the anode choke, now 27H, and 273 ohm (+/- like a good output transformer) This thing has almost no influence on results, for now.

I'm writing a mail to Langrex, through ebay, in order to inform the seller about the problem.

[PJs]

-initial circuit run at 55v between gnd and cathode, with 220ohm cathode resistor are 250ma current, 3 times and something over normal operation. Not good, this working point destroyed first resistor

Yup, a 5W resistor won't hold up to 13.75 watts very long...surprised nothing else went. Can't understand how a matched tube can be that different unless the cathode was shorted internally somehow¿ And the Bakelite base of that tube is singed! @!@ Hopefully your power supply is OK and they will send you a replacement?

I'm assuming this is the same "Cathode Follower" circuit you showed me before. That's not a typical setup for a CF...Might want to take a look at that as I suggested before. Think I sent some links on them way back in the thread...if not Google it.

~PJ

[rendoman]

Yes sir! Circuit is similar, updated for different tubes, in order to have more or less the same power of oscillator part, a good old rule with old transmitter, as I read

CF part is good as in the past,luckily something works! With 300v anode supply I get half voltage output to drive oscillator.
Not enough to bring the hell on earth, old good full flame with bright red plate near 200ma 1,1kv .
Still enough to achive a big, not extreme flame with still grey plate. I will play with coils and working frequency, even if I think that actual freq is good. Although I must say that this circuit performs better with tubes really hot.

The most spectacular working point tried, unfortunately I got no pictures (and maybe I never wrote here, because it was crazy) , was 1,3kv with ammeter in heavy over range, up to 250ma, with the typical sound of pointer against block. Small flame and pesky bright white from anode... That's why tubes hate me

I hope ebay seller will reply Tomorrow, I think as you, some short circuit inside, or some strange grids structure, no signs on bakelite base, or visibile inside, the trouble should be inside.
I hope for a replacement, even if I bought a quad matched set ( I don't believe ), and I wonder how they can add a single tube with no matching reference.... I think I should close both eyes

[rendoman]

Hi all!
I tried some circuit with fun tonight.
Because of many trouble with circuit, I wanted to test single el34 tube. I decided to make a fast amplifier on a relay socket, driven by old audio generator, capable of delivering 10v rms at 1khz, according to datasheet enough for undistorted output. El34 in triode now is good, but I have for the second time sockets-parts issues. Hope to find the cause soon

I will try in the coming days different bias and pentode mode, plus a better driver. Now tube is running at 52ma, with light warm cathode resistor, I'm on the right path

I also measured my ecc82 first stage, 35v rms output with less than 0,2v rms input, with a nice sine.

For the trouble with seller, he sent me a spare valve, which it is , more or less, a confirmation that there is no a real matching.



Ecc82 driver, last picture with 227khz

[PJs]

Hi Stefano, I'm not sure what I'm seeing on the scope...assume channel 1 is the EL34 and channel 2 is the ECC82 signal outputs¿

Is this the spare tube? We both new they weren't matching these to any real spec...its a huge deal to go through all of a production run and matching tube curves plus aging times and precision setups. That is precisely why they Cost so much. This was just Advertising to suck you in.

Those old 833's I had were matched by RCA and went for about $300 a pair back in the day!! These were on Ebay and don't even say matched pair Nor have gold Grids and definitely are Not RCA! But here is a real find on Ebay.

The sine waves look Very good at 1khz, spec voltages, but found it interesting that at 18khz channel 2 leads channel 1 and at 20khz it lags a bit? It appears that pushing the voltage on channel 2 is creating a phase shift...maybe slightly over driven. The good thing is even at 228KHz all you are getting is a ~90º phase shift and no distortion...that is impressive...be curious to see if it comes back to normal with a 35v on the ECC82.

You seem to be on a good path now, Congrats!

Until then, ~~ Keep your voltages nominal, your gain to 85-90% and your eye on the prize! Sparks will fly.
~PJ

[rendoman]

Hi!
the tube is still traveling, Hope I will have in few days. You're right, pure advertising about matching, like many Others dealer. Next step on metal frame, I have buy belton, or anyway new high quality sockets, not chinese, something good, in order to have at least some more quality.
833a, love this tube!

Ch1 is Always linked to signal generator, ch2 on output of el34 in first set of pictures, and output of ecc82 in second one. You have a very good eye! I noticed that at 18khz output sine appears a bit translated, I will do a better thorough test to see correct levels at diferent sample rate. It curious too the 227khz test, I did for pure curiosity, but I did not expect this results

Thanks for the advice! coming soon new tests!

ps: this is the chinese datasheet of el34 tube. Ecc82 is a 6189w special quality, siemens

[rendoman]

Hi!
Today I received the single spare el34 from Langrex, for free, it's a good seller for this.
I tried the faulty tube in new circuit, still gives bad results, I think I will test my new glock (arrival scheduled for this week) at shooting range with tube as target

I made some experiments with ecc82, single ended, one or double triodes connected. Both circuit give good results, for a se amp with el34, a single unit should be fine.

With 1v rms input (2.83v p-p), I got 13.7v rms or 38.8v p-p, enough to drive el34, maybe the right choice.

I tried el34 triode at 52 and 78ma, with same B+ but different cathode resistor. Unfortunately I can't find ceramic 8 ohm resistor as dummy load, I only see the anode amplitude for now. 52ma gives a better swing, and a colder rk, 78ma give imho too high temperature on 10w resistor and less amplitude. I will try soon with 8 ohm load and another output transformer

I tried also el34 in triode mode with good results, I think I will build an amplifier for my setup, this tube worth!



Ecc82 single ended, half triode


El34 triode, 19,5v bias, 375ohm cathode resistor


El34 triode, 16v bias, 220ohm cathode resistor

[PJs]

Looking Good Stefano! Glad they comp'd the tube for you!

The 16v/220 KR seems to lag just a skosh but the 19.5v/375 KR leads a bit more than a skosh...might try a 330 ohm at the 16v. That should put it at 0º...and the amplitude looks better at those lower values. It should be plenty of gain for your need but will depend how well you couple it and the quality of the oscillator/HF circuit.

Not sure what to say about the 8/10 ohm...more henry's might help???

Keep up the faith and the force will be with you! Great work!

~PJ

[rendoman]

Hi all!
Big update this time!

After a night, my plasma speaker is running!
Trouble as usual:* 2 new capacitor purchased from an Italian shop (the same of noval defective sockets).*

Few words about, I did a fast run, no hum, no noise, good and loud sound. The shielded frame works perfectly even if I soldered fast some components inside.

I have to try for my curiosity the range with microphone and software, 1khz is loud, maybe 2khz is the usable fq.*Finally I can tune the circuit, there is lot of work to do, tubes run in full safe operation mode now, I can squeeze more power from both*hf unit and modulator (remaining in datasheet range).

Next attempts:

- pentode mode amplifier unit
- raise driving voltage of oscillator at maximum (datasheet)
- try higher mhz working frequency, different coils and cap (flame is fully silent)

With my new modulator circuit flame runs from 10 to 20mm, from 50 to 110ma at 1kv.
No red plate, no red grids, power transformers run at low temperature.
Autofire is Always running, no strange behavior noticed.
In few words: circuit is stable and well running, now is labor limae!

For comfort I used semi-faraday gage, it works perfectly, see differences in pictures with neon, at the side is not possible to fire neon, even the rf emission on top is really reduced compared to full air coil.